Issue #271 - Allow missing values when an sdmxa:obsStatus Attribute explains why it's missing.
#272
Conversation
… explains why it's missing.
csvcubed/csvcubed/utils/qb/cube.py
Outdated
| return errors | ||
|
|
||
|
|
||
| def _validate_missing_observation_values( |
There was a problem hiding this comment.
I think that the observations should be in their own module.
There was a problem hiding this comment.
I think at some point we'll end up splitting out all of the different validations into separate files. What makes you suggest it now?
There was a problem hiding this comment.
It's grown this module by 50% LoC, which is some artibtrary mental threshold for some minor refactoring. Won't hurt if you don't want to do it now, but it is easily done now.
| # Unfortunately, CSV-W validation will *not* catch this error since the obs column cannot be marked as `required` | ||
| # since an `sdmxa:obsStatus` Attribute column has been defined. | ||
| Then csvlint validation of "bad-qube.csv-metadata.json" should succeed |
There was a problem hiding this comment.
Is there a thing to do with csvcheck that could solve this?
There was a problem hiding this comment.
I don't think csvw-check is the right place to solve the issue. We'd effectively be making an unofficial extension to their CSV-W spec.
However, there is definitely scope for more CSV-qb validation which I believe is in our issues long-term.
csvcubed/csvcubed/utils/qb/cube.py
Outdated
| return errors | ||
|
|
||
|
|
||
| def _validate_missing_observation_values( |
There was a problem hiding this comment.
It's grown this module by 50% LoC, which is some artibtrary mental threshold for some minor refactoring. Won't hurt if you don't want to do it now, but it is easily done now.
Issues raised as a result of this PR:
Satisfies Issue #271.